Resolves: fdo#45271, i#17964 count CJK words the way that's expected by users

6 files changed, 124 insertions, 10 deletions

diff --git a/sw/inc/docstat.hxx b/sw/inc/docstat.hxx
index 554d506be414..1869120b15ee 100644
--- a/sw/inc/docstat.hxx
+++ b/sw/inc/docstat.hxx
@@ -43,6 +43,7 @@ struct SW_DLLPUBLIC SwDocStat
     /// all paragraphs, including empty/hidden ones
     sal_uLong           nAllPara;
     sal_uLong           nWord;
+    sal_uLong           nAsianWord;
     sal_uLong           nChar;
     sal_uLong           nCharExcludingSpaces;
     sal_Bool            bModified;
diff --git a/sw/inc/ndtxt.hxx b/sw/inc/ndtxt.hxx
index 21e9fb4aa71e..83b3e2f35a4a 100644
--- a/sw/inc/ndtxt.hxx
+++ b/sw/inc/ndtxt.hxx
@@ -173,6 +173,8 @@ class SW_DLLPUBLIC SwTxtNode: public SwCntntNode, public ::sfx2::Metadatable
     //
     SW_DLLPRIVATE void SetParaNumberOfWords( sal_uLong nTmpWords ) const;
     SW_DLLPRIVATE sal_uLong GetParaNumberOfWords() const;
+    SW_DLLPRIVATE void SetParaNumberOfAsianWords( sal_uLong nTmpAsianWords ) const;
+    SW_DLLPRIVATE sal_uLong GetParaNumberOfAsianWords() const;
     SW_DLLPRIVATE void SetParaNumberOfChars( sal_uLong nTmpChars ) const;
     SW_DLLPRIVATE sal_uLong GetParaNumberOfChars() const;
     SW_DLLPRIVATE void SetParaNumberOfCharsExcludingSpaces( sal_uLong nTmpChars ) const;
diff --git a/sw/qa/core/swdoc-test.cxx b/sw/qa/core/swdoc-test.cxx
index 72c5ad32e276..8469c0d38c7c 100644
--- a/sw/qa/core/swdoc-test.cxx
+++ b/sw/qa/core/swdoc-test.cxx
@@ -113,7 +113,7 @@ void SwDocTest::testPageDescName()
     CPPUNIT_ASSERT_MESSAGE("GetPageDescName results must be unique", aResults.size() == 3);
 }
 
-//See https://bugs.freedesktop.org/show_bug.cgi?id=32463 for motivation
+//See https://bugs.freedesktop.org/show_bug.cgi?id=32463
 void SwDocTest::testFileNameFields()
 {
     //Here's a file name with some chars in it that will be %% encoded, when expanding
@@ -225,8 +225,8 @@ void SwDocTest::testSwScanner()
 
     CPPUNIT_ASSERT_MESSAGE("Has Text Node", pTxtNode);
 
-    //See https://bugs.freedesktop.org/show_bug.cgi?id=40449 for motivation
-    //See https://bugs.freedesktop.org/show_bug.cgi?id=39365 for motivation
+    //See https://bugs.freedesktop.org/show_bug.cgi?id=40449
+    //See https://bugs.freedesktop.org/show_bug.cgi?id=39365
     //Use a temporary rtl::OUString as the arg, as that's the trouble behind
     //fdo#40449 and fdo#39365
     {
@@ -248,7 +248,7 @@ void SwDocTest::testSwScanner()
             rWorld.equalsAsciiL(RTL_CONSTASCII_STRINGPARAM("World")));
     }
 
-    //See https://www.libreoffice.org/bugzilla/show_bug.cgi?id=45271 for motivation
+    //See https://www.libreoffice.org/bugzilla/show_bug.cgi?id=45271
     {
         const sal_Unicode IDEOGRAPHICFULLSTOP_D[] = { 0x3002, 'D' };
 
@@ -261,13 +261,51 @@ void SwDocTest::testSwScanner()
         m_pDoc->InsertPoolItem(aPaM, aWestLangItem, 0 );
 
         SwDocStat aDocStat;
+        pTxtNode = aPaM.GetNode()->GetTxtNode();
         pTxtNode->CountWords(aDocStat, 0, SAL_N_ELEMENTS(IDEOGRAPHICFULLSTOP_D));
 
         CPPUNIT_ASSERT_MESSAGE("Should be 2", aDocStat.nChar == 2);
         CPPUNIT_ASSERT_MESSAGE("Should be 2", aDocStat.nCharExcludingSpaces == 2);
     }
+    {
+        const sal_Unicode test[] =
+        {
+            0x3053, 0x306E, 0x65E5, 0x672C, 0x8A9E, 0x306F, 0x6B63, 0x3057,
+            0x304F, 0x6570, 0x3048, 0x3089, 0x308C, 0x308B, 0x3067, 0x3057,
+            0x3087, 0x3046, 0x304B, 0x3002, 0x0041, 0x006E, 0x0064, 0x0020,
+            0x006C, 0x0065, 0x0074, 0x0027, 0x0073, 0x0020, 0x0074, 0x0068,
+            0x0072, 0x006F, 0x0077, 0x0020, 0x0073, 0x006F, 0x006D, 0x0065,
+            0x0020, 0x0045, 0x006E, 0x0067, 0x006C, 0x0069, 0x0073, 0x0068,
+            0x0020, 0x0069, 0x006E, 0x0020, 0x0074, 0x006F, 0x0020, 0x006D,
+            0x0061, 0x006B, 0x0065, 0x0020, 0x0069, 0x0074, 0x0020, 0x0069,
+            0x006E, 0x0074, 0x0065, 0x0072, 0x0065, 0x0073, 0x0074, 0x0069,
+            0x006E, 0x0067, 0x002E, 0x0020, 0x0020, 0x305D, 0x3057, 0x3066,
+            0x3001, 0x307E, 0x305F, 0x65E5, 0x672C, 0x8A9E, 0x3000, 0x3000,
+            0x3067, 0x3082, 0x4ECA, 0x56DE, 0x306F, 0x7A7A, 0x767D, 0x3092,
+            0x3000, 0x3000, 0x5165, 0x308C, 0x307E, 0x3057, 0x305F, 0x3002,
+            0x0020, 0x0020, 0x0053, 0x006F, 0x0020, 0x0068, 0x006F, 0x0077,
+            0x0020, 0x0064, 0x006F, 0x0065, 0x0073, 0x0020, 0x0074, 0x0068,
+            0x0069, 0x0073, 0x0020, 0x0064, 0x006F, 0x003F, 0x0020, 0x0020
+        };
+        m_pDoc->AppendTxtNode(*aPaM.GetPoint());
+        m_pDoc->InsertString(aPaM, rtl::OUString(test,
+            SAL_N_ELEMENTS(test)));
+
+        SvxLanguageItem aCJKLangItem( LANGUAGE_JAPANESE, RES_CHRATR_CJK_LANGUAGE );
+        SvxLanguageItem aWestLangItem( LANGUAGE_ENGLISH_US, RES_CHRATR_LANGUAGE );
+        m_pDoc->InsertPoolItem(aPaM, aCJKLangItem, 0 );
+        m_pDoc->InsertPoolItem(aPaM, aWestLangItem, 0 );
+
+        SwDocStat aDocStat;
+        pTxtNode = aPaM.GetNode()->GetTxtNode();
+        pTxtNode->CountWords(aDocStat, 0, SAL_N_ELEMENTS(test));
+        CPPUNIT_ASSERT_MESSAGE("58 words", aDocStat.nWord == 58);
+        CPPUNIT_ASSERT_MESSAGE("43 Asian characters and Korean syllables", aDocStat.nAsianWord == 43);
+        CPPUNIT_ASSERT_MESSAGE("105 non-whitespace chars", aDocStat.nCharExcludingSpaces == 105);
+        CPPUNIT_ASSERT_MESSAGE("128 characters", aDocStat.nChar == 128);
+    }
 
-    //See https://issues.apache.org/ooo/show_bug.cgi?id=89042 for motivation
+    //See https://issues.apache.org/ooo/show_bug.cgi?id=89042
     {
         SwDocStat aDocStat;
 
@@ -298,8 +336,7 @@ void SwDocTest::testSwScanner()
     }
 }
 
-
-//See https://bugs.freedesktop.org/show_bug.cgi?id=40599 for motivation
+//See https://bugs.freedesktop.org/show_bug.cgi?id=40599
 void SwDocTest::testGraphicAnchorDeletion()
 {
     CPPUNIT_ASSERT_MESSAGE("Expected initial 0 count", m_pDoc->GetDocStat().nChar == 0);
diff --git a/sw/source/core/doc/docstat.cxx b/sw/source/core/doc/docstat.cxx
index c84f943b8516..fd0c6e65aa57 100644
--- a/sw/source/core/doc/docstat.cxx
+++ b/sw/source/core/doc/docstat.cxx
@@ -43,6 +43,7 @@ SwDocStat::SwDocStat() :
     nPara(1),
     nAllPara(1),
     nWord(0),
+    nAsianWord(0),
     nChar(0),
     nCharExcludingSpaces(0),
     bModified(sal_True)
@@ -61,6 +62,7 @@ void SwDocStat::Reset()
     nPara   = 1;
     nAllPara= 1;
     nWord   = 0;
+    nAsianWord = 0;
     nChar   = 0;
     nCharExcludingSpaces = 0;
     bModified = sal_True;
diff --git a/sw/source/core/txtnode/txtedt.cxx b/sw/source/core/txtnode/txtedt.cxx
index a186cc646857..cb5e4724fe8c 100644
--- a/sw/source/core/txtnode/txtedt.cxx
+++ b/sw/source/core/txtnode/txtedt.cxx
@@ -680,6 +680,44 @@ SwScanner::SwScanner( const SwTxtNode& rNd, const rtl::OUString& rTxt,
     }
 }
 
+namespace
+{
+    //fdo#45271 for Asian words count characters instead of words
+    sal_Int32 forceEachAsianCodePointToWord(const rtl::OUString &rText, sal_Int32 nBegin, sal_Int32 nLen)
+    {
+        if (nLen > 1)
+        {
+            const uno::Reference< XBreakIterator > &rxBreak = pBreakIt->GetBreakIter();
+
+            sal_uInt16 nCurrScript = rxBreak->getScriptType( rText, nBegin );
+
+            sal_Int32 indexUtf16 = nBegin;
+            rText.iterateCodePoints(&indexUtf16, 1);
+
+            //First character is Asian, consider it a word :-(
+            if (nCurrScript == i18n::ScriptType::ASIAN)
+            {
+                nLen = indexUtf16 - nBegin;
+                return nLen;
+            }
+
+            //First character was not Asian, consider appearance of any Asian character
+            //to be the end of the word
+            while (indexUtf16 < nBegin + nLen)
+            {
+                nCurrScript = rxBreak->getScriptType( rText, indexUtf16 );
+                if (nCurrScript == i18n::ScriptType::ASIAN)
+                {
+                    nLen = indexUtf16 - nBegin;
+                    return nLen;
+                }
+                rText.iterateCodePoints(&indexUtf16, 1);
+            }
+        }
+        return nLen;
+    }
+}
+
 sal_Bool SwScanner::NextWord()
 {
     nBegin = nBegin + nLen;
@@ -802,6 +840,9 @@ sal_Bool SwScanner::NextWord()
     if( ! nLen )
         return sal_False;
 
+    if ( nWordType == i18n::WordType::WORD_COUNT )
+        nLen = forceEachAsianCodePointToWord(aText, nBegin, nLen);
+
     aWord = aText.copy( nBegin, nLen );
 
     return sal_True;
@@ -1812,6 +1853,7 @@ void SwTxtNode::CountWords( SwDocStat& rStat,
     {
         // accumulate into DocStat record to return the values
         rStat.nWord += GetParaNumberOfWords();
+        rStat.nAsianWord += GetParaNumberOfAsianWords();
         rStat.nChar += GetParaNumberOfChars();
         rStat.nCharExcludingSpaces += GetParaNumberOfCharsExcludingSpaces();
         return;
@@ -1842,7 +1884,8 @@ void SwTxtNode::CountWords( SwDocStat& rStat,
     // all counts exclude hidden paras and hidden+redlined within para
     // definition of space/white chars in SwScanner (and BreakIter!)
     // uses both lcl_IsSkippableWhiteSpace and BreakIter getWordBoundary in SwScanner
-    sal_uInt32 nTmpWords = 0;        // count of all contiguous blocks of non-white chars
+    sal_uInt32 nTmpWords = 0;        // count of all words
+    sal_uInt32 nTmpAsianWords = 0;   //count of all Asian codepoints
     sal_uInt32 nTmpChars = 0;        // count of all chars
     sal_uInt32 nTmpCharsExcludingSpaces = 0;  // all non-white chars
 
@@ -1862,7 +1905,10 @@ void SwTxtNode::CountWords( SwDocStat& rStat,
             if( 1 != aExpandText.match(aBreakWord, aScanner.GetBegin() ))
             {
                 ++nTmpWords;
-                nTmpCharsExcludingSpaces += pBreakIt->getGraphemeCount(aScanner.GetWord());
+                const rtl::OUString &rWord = aScanner.GetWord();
+                if (pBreakIt->GetBreakIter()->getScriptType(rWord, 0) == i18n::ScriptType::ASIAN)
+                    ++nTmpAsianWords;
+                nTmpCharsExcludingSpaces += pBreakIt->getGraphemeCount(rWord);
             }
         }
     }
@@ -1890,7 +1936,10 @@ void SwTxtNode::CountWords( SwDocStat& rStat,
             while ( aScanner.NextWord() )
             {
                 ++nTmpWords;
-                nTmpCharsExcludingSpaces += pBreakIt->getGraphemeCount(aScanner.GetWord());
+                const rtl::OUString &rWord = aScanner.GetWord();
+                if (pBreakIt->GetBreakIter()->getScriptType(rWord, 0) == i18n::ScriptType::ASIAN)
+                    ++nTmpAsianWords;
+                nTmpCharsExcludingSpaces += pBreakIt->getGraphemeCount(rWord);
             }
 
             nTmpChars = pBreakIt->getGraphemeCount(aNumString);
@@ -1909,12 +1958,14 @@ void SwTxtNode::CountWords( SwDocStat& rStat,
     if ( isCountAll )
     {
         SetParaNumberOfWords( nTmpWords );
+        SetParaNumberOfAsianWords( nTmpAsianWords );
         SetParaNumberOfChars( nTmpChars );
         SetParaNumberOfCharsExcludingSpaces( nTmpCharsExcludingSpaces );
         SetWordCountDirty( false );
     }
     // accumulate into DocStat record to return the values
     rStat.nWord += nTmpWords;
+    rStat.nAsianWord += nTmpAsianWords;
     rStat.nChar += nTmpChars;
     rStat.nCharExcludingSpaces += nTmpCharsExcludingSpaces;
 }
@@ -1928,6 +1979,7 @@ struct SwParaIdleData_Impl
     SwGrammarMarkUp* pGrammarCheck;     // for grammar checking /  proof reading
     SwWrongList* pSmartTags;
     sal_uLong nNumberOfWords;
+    sal_uLong nNumberOfAsianWords;
     sal_uLong nNumberOfChars;
     sal_uLong nNumberOfCharsExcludingSpaces;
     bool bWordCountDirty;
@@ -1941,6 +1993,7 @@ struct SwParaIdleData_Impl
         pGrammarCheck       ( 0 ),
         pSmartTags          ( 0 ),
         nNumberOfWords      ( 0 ),
+        nNumberOfAsianWords ( 0 ),
         nNumberOfChars      ( 0 ),
         nNumberOfCharsExcludingSpaces ( 0 ),
         bWordCountDirty     ( true ),
@@ -2033,10 +2086,25 @@ void SwTxtNode::SetParaNumberOfWords( sal_uLong nNew ) const
         m_pParaIdleData_Impl->nNumberOfWords = nNew;
     }
 }
+
 sal_uLong SwTxtNode::GetParaNumberOfWords() const
 {
     return m_pParaIdleData_Impl ? m_pParaIdleData_Impl->nNumberOfWords : 0;
 }
+
+void SwTxtNode::SetParaNumberOfAsianWords( sal_uLong nNew ) const
+{
+    if ( m_pParaIdleData_Impl )
+    {
+        m_pParaIdleData_Impl->nNumberOfAsianWords = nNew;
+    }
+}
+
+sal_uLong SwTxtNode::GetParaNumberOfAsianWords() const
+{
+    return m_pParaIdleData_Impl ? m_pParaIdleData_Impl->nNumberOfAsianWords : 0;
+}
+
 void SwTxtNode::SetParaNumberOfChars( sal_uLong nNew ) const
 {
     if ( m_pParaIdleData_Impl )
@@ -2044,10 +2112,12 @@ void SwTxtNode::SetParaNumberOfChars( sal_uLong nNew ) const
         m_pParaIdleData_Impl->nNumberOfChars = nNew;
     }
 }
+
 sal_uLong SwTxtNode::GetParaNumberOfChars() const
 {
     return m_pParaIdleData_Impl ? m_pParaIdleData_Impl->nNumberOfChars : 0;
 }
+
 void SwTxtNode::SetWordCountDirty( bool bNew ) const
 {
     if ( m_pParaIdleData_Impl )
diff --git a/sw/source/ui/dialog/wordcountdialog.cxx b/sw/source/ui/dialog/wordcountdialog.cxx
index 3fdc8958ddc0..f5b1c3b98824 100644
--- a/sw/source/ui/dialog/wordcountdialog.cxx
+++ b/sw/source/ui/dialog/wordcountdialog.cxx
@@ -42,6 +42,8 @@
 #include <swwait.hxx>
 #include <wrtsh.hxx>
 
+//TODO, add asian/non-asian word count to UI when CJK mode is enabled.
+
 SwWordCountDialog::SwWordCountDialog(Window* pParent) :
 #if defined _MSC_VER
 #pragma warning (disable : 4355)